Individual-Focused College Student Drinking Prevention:
Revisiting the 2002 NIAAA Task Force Report
Jessica M. Cronce, Ph.D.
Center for the Study of Health & Risk Behaviors
Department of Psychiatry & Behavioral Sciences
University of Washington
Campus reactions to the
2002 NIAAA Task Force report
4
5
Disagree/Other
Reviewed but
not yet
implemented
Implemented
some of them
8
Waiting to be
mailed
16
3
In the process
of reading
39
21
Implemented
all of them
100
90
80
70
60
50
40
30
20
10
0
Not aware
%
Awareness of the 2002 NIAAA Task Force Recommendations
Source: Nelson et al. (2010)
Campus reactions to the
2002 NIAAA Task Force report
76% of colleges surveyed offered 1 or more Tier 1 program
%
Tier 1 Programs Implemented
100
90
80
70
60
50
40
30
20
10
0
66
63
57
38
Norms
clarification
Cognitivebehavioral skills
training
Motivational
interviewing
Expectancy
challenge
Source: Nelson et al. (2010)
Plenary Goals
…
Increase awareness of the NIAAA College Drinking
Task Force’s tier system of efficacy &
recommended strategies
…
Review the body of evidence that supported the
Task Force recommendations
…
Enhance understanding of commonalities and
differences among recommended “Tier 1” programs
…
Share “hot off the presses” evidence that
updates the Task Force report, including new
programs
…
Summarize key “take home” messages
…
Leave 5-10 minutes for Q&A
NIAAA Task Force Tier System
…
…
…
…
Tier I: Evidence of effectiveness
among college students (≥2
studies supporting efficacy)
Tier 2: Evidence of success with
other populations that could be
applied to college environments
Tier 3: Evidence of logical and
theoretical promise, but require
more comprehensive evaluation
Tier 4: Evidence of ineffectiveness
www.CollegeDrinkingPrevention.gov
# of Studies
Evidence considered in the
NIAAA Task Force Recommendations
10
9
8
7
6
5
4
3
2
1
0
1
7
8
2
1
3
MCST
BMI
AEC
“Tier 1”
Positive
Effects
6
Null/Neg.
Effects
Education
Only
MCST: Combining cognitive-behavioral skills with norms clarification and
motivational enhancement; a.k.a., multi-component skills training
BMI:
Brief motivational enhancement interventions
AEC:
Alcohol expectancy challenge
What’s COMMON to Tier 1 Strategies?
1. Teach moderatedrinking & life
management skills
3. Enhance motivation
for change
5. Challenge positive
alcohol expectancies
2. Alcohol education
to support skill-use
4. Correct misperceived
drinking norms
Drinks per occasion
16
14
12
10
8
6
4
2
0
1 container
≠
You
drink…
You
Other
thought students
other
actually
students
drink...
drink…
1 drink
Alcohol does NOT
make you
“Dancing with the
Stars” material.
What DIFFERENTIATES Tier 1 strategies?
# of
Sessions
Focus
Building/strengthening
MCST safer-drinking skills
BMI
Increasing awareness
& motivation for change
AEC
Challenging positive
expectancies
Structure
Guided by
4 to 6+
In-person Group
Manual /
agenda
1 or 2
In-person Individual
(or Group)*
Personalized
feedback*
In-person Group
Alcohol
administration
1
Personalized feedback is a cornerstone of the
BASICS program, the most well-evaluated BMI.
Delivered as a stand-alone print or electronic
intervention:
PFI = personalized feedback intervention
PNF = personalized normative feedback
PFI & PNF
BMI > PFI > PNF
Examples of
Web-based PFI Î
Exponential growth in alcohol prevention RCTs
Larimer & Cronce
(2007): 42 studies
Cronce & Larimer
(2011)*: 36 studies
15
S tu dies per year
Larimer & Cronce
(2002): 32 studies
10
5
0
1984 late 1999 2007 early 1999
2006
early 2010
*Alcohol Research & Health, 34(2), 210-221.
http://pubs.niaaa.nih.gov/publications/arh342/210-221.pdf
Summarizing the Evidence
“Thumbs up” =
Reduced drinking and/or
related consequences; OR
protective effect
“Thumbs down” =
Intervention no different
than assessment alone; OR
increased drinking
MCST (CBT skills, Norms Clarification & MET)
&
'
# Evaluated
1984-2010
10
8
6
4
2
0
7
3
4
4
2
13
9
2
2002 2007 2011
ƒ
Apparent trend toward fewer studies evaluating MCST, perhaps
due to increased resource/participant burden relative to BMI
ƒ
Parent-based intervention (PBI) facilitating communication
around alcohol use combined with BMI for students is more
effective than BMI alone in preventing negative consequences
In-person BMI (most with PFI/PNF)
1984-2010
# Evaluated
20
&
'
6
15
10
5
10
8
4
0
17
35
2
2002 2007 2011
ƒ
ƒ
ƒ
Effects on negative consequences may persist up to 4 years (BASICS)
Potential delayed effects on negative consequences
BMI-alone and MCST-alone are equivalent on most outcomes, but BMI
may be superior to MCST for some outcomes (3 studies: negative
consequences; high-risk drinking; weekend/weekday quantity)
AEC interventions
1984-2010
&
'
# Evaluated
8
2
6
2
4
2
0
2
1
5
2
2002 2007 2011
ƒ
Experiential
4
Didactic/video
Experiential AEC shown to be effective with men.
ƒ
ƒ
6
Mixed findings for women, with some evidence of positive effects when
gender-specific expectancies are challenged in single-gender groups.
Didactic and video AEC generally not effective
&
'
# Evaluated
Education/Awareness ONLY
1984-2010
10
8
6
4
2
0
1
2
1
9
6
4
19
2002 2007 2011
ƒ
Education/awareness only continues to be ineffective in changing
drinking outcomes other than alcohol knowledge
ƒ
Many studies included an education condition as comparison group
ƒ
Only 1 new study since 2007 that evaluated education-only
Stand-alone computerized or mailed
PFI (most with PNF) 1984-2010
&
'
# Evaluated
15
10
5
7
3
0
2002
ƒ
ƒ
ƒ
ƒ
5
10
1
2007
4
20
2011
Few studies compare BMI w/ PFI to PFI-alone (only 6 since 1999)
BMI w/ PFI and PFI-alone comparable on most outcomes
2 found in-person BMI w/ PFI to be more efficacious than stand-alone
PFI on at least some outcomes (e.g., drinking/consequences composite)
Total includes 3 evaluations of e-Check Up to Go (e-CHUG) with
positive results (decreased drinking [3] & consequences [1])
Stand-alone PNF (incl. ESP)
1999-2010
&
'
# Evaluated
4
3
4
2
1
1
3
7
1
0
2002
2007
2011
ƒ
Changes in norms mediated effects on drinking outcomes
ƒ
Level of specificity with respect to reference group may influence
outcomes (e.g., 1 study found gender-specific PNF more effective than
gender-neutral PNF)
ƒ
Findings for event specific prevention (ESP) related to 21st birthday
drinking outcomes are mixed.
Multi-component education-focused programs
1999-2010
&
'
?
ƒ
ƒ
ƒ
# Evaluated
8
2
4
6
4
6
2
4
2
4
2
0
2002
2007
2011
Programs that have historically been education-only, or predominant education
component, but include some elements found in BMI / PFI / PNF.
Efficacy may be version specific, and conclusions should NOT be generalized.
MCEFPs include: Alcohol 101 (-2, ~2), Alcohol 101 Plus (-1, +1),
College Alc (+1), MyStudentBody.com (+1), AlcoholEdu for Sanctions
(+1), AlcoholEdu for College (3…)
AlcoholEdu for College RCTs (in Cronce & Larimer, 2011)
&
&
'
Hustad et al., 2010 (version 9.0 [per Wyatt, DeJong & Dixon, in press]; only
included 18+)
ƒ Decreased alcohol consumption (or smaller increases in alcohol
consumption) equal to e-CHUG and > assessment only (AO).
ƒ Decreased negative consequences > AO and no different than e-CHUG,
although decreases in e-CHUG were statistically ns.
Lovecchio et al., 2010 (version 8.0; only included 18+)
Smaller increases in alcohol consumption > AO .
ƒ Decreased negative consequences & positive alcohol expectancies > AO.
ƒ No differences on high-risk or protective alcohol behaviors.
ƒ Decreases in responsible drinking behaviors.
ƒ
Croom et al., 2009 (2004, 2006 or 2007 version?; included 17+)
ƒ Both groups increased alcohol consumption, consequences, and other
alcohol-related risk behaviors, with no significant differences between
groups.
ƒ Exception: Smaller % played drinking games, but larger % failed to use
safe sex practices.
Other studies of AlcoholEdu:
Wall, 2007 (2003 version?; ages?)
ƒ
?
ƒ
ƒ
ƒ
Did not randomly assign individuals (or groups) to intervention or control.
“Randomization” was post hoc, comparing “control participants’” pre-test scores
to “intervention participants’” post-test scores, which doesn’t control for the effect
of assessment reactivity.
Extremely OVER powered (N = 20,150), so differences questionable.
Immediate post-test only with no follow-up.
Paschall et al., 2011 (version not specified; only included 18+)
ƒ
ƒ
&
?
ƒ
ƒ
ƒ
ƒ
Random assignment at the level of the university.
New cross-sectional random sample (N=200) of students each quarter; 44%49% survey response rate.
Wide range of intervention completion rate: 4%-100%.
Decreased frequency of alcohol use and binge drinking relative to control at
immediate post-assessment (Fall).
No effect evident in Spring, even among schools with high intervention
completion rates.
Did not examine alcohol-related consequences.
Wyatt, DeJong & Dixon, in press – Time series analysis (not RCT)
Special Populations: Mandated Students
&
'
# Evaluated
1984-2010
9
8
7
6
5
4
3
2
1
0
8
3
4
3
1
1
1
MCST
BMI
AEC
1
2
1
PFI/PNF MCEFP Education
Only
Pros/cons of different prevention programs
MCST
(e.g., ASTP)
BMI (e.g.,
BASICS)
PFI / PNF
(e.g., e-CHUG)
Experiential
AEC
AlcoholEdu
Other MCEFP
(e.g.,
CollegeAlc)
Development/
Training Cost
$
$$
*$ - $$
-
-
-
Implementatio
n Cost
$$
$$
*$ - $$
$$$
$$$$
$$ - $$$$
1-2 people
1-2 people
-
1-2 people
-
-
Specialized
Training Req.
Yes
Yes
No
Yes
No
No
Specialized
Resources
Req.
No
No
No
Yes
Yes
Yes
Reach
Groups (8-12
students)
Individual
students
All students
Groups (8-12
students)
All
students
All students
Range of
Effect Sizes**
d = 0.13-0.26
(w/ BMI)
d = 0.211.06
d = 0.29-0.85
d = 0.00-0.36
d = 0.560.75
d = 0.15-0.38
Length of
Effects on
Drinking
Up to 1 year
Up to 4
years
Up to 1 year
Up to 3
months
Up to 1
month
Variable (shortterm)
Human
Resources
* FREE PFI: Check Your Drinking beta version: http://notes.camh.net/efeed.nsf/newform
Doumas & Hannah, 2008; Doumas & Haustveit, 2008; Doumas, McKinley & Book, 2009 all found positive effects of this version.
** Based on studies included in Cronce & Larimer, 2011
Conclusions: Looking BACK
…
…
Overwhelming support for BMI and related interventions
incorporating Motivational Interviewing style, PFI, PNF,
and AEC components
†
Evidence supporting e-CHUG, CheckYourDrinking.net and
other electronic personalized feedback programs adds to
growing evidence for PFIs.
†
Less consistency on changing consequences than drinking
per se
†
Longer follow-ups necessary; in-person BMI associated with
emergent effects on consequences
Emerging evidence in support of one MCEFP—
AlcoholEdu for College—but more research is needed
(RCTs = 4 vs. 41 BMI, 25 PFI).
Conclusions: Looking FORWARD
…
…
Need more research on BMIs and PFIs targeting multiple risk
behaviors & spanning the alcohol/mental health divide
†
Alcohol and marijuana use
†
Alcohol/marijuana use and problematic gambling behavior
†
Depression and alcohol use
Future RCTs of AlcoholEdu, e-CHUG and other electronic
prevention programs would benefit from:
†
Including matriculating freshman <18 years old.
†
Incorporating baseline & post-assessment that is independent of the
program.
†
Gauging level of engagement and depth of processing.
†
Including longer, longitudinal follow-up.
Take Home Messages
…
Individual-focused programs need to be considered one
“piece” of a larger “prevention puzzle.”
…
Strategies recommended by the NIAAA Task Force (i.e.,
MCST, BMI, experiential AEC) continue to produce
drinking reductions, but there are other options that
current science suggests work (i.e., PFIs, PNF)
…
Reach needs to be weighed against strength and
duration of effect taking into consideration initial/ongoing
costs and resource demands.
†
…
Some costs can be diffused through collaboration &
technology
The science is constantly evolving, and prevention
approaches need to be regularly revisited.
Thank You!
and thank you to…
NIAAA (T32 AA007455)
Mary E. Larimer, Ph.D.
Jason R. Kilmer, Ph.D.
Stephanie Gordon